A variety of service nodes are generally deployed in a service center or a network edge point of presence (PoP) provided by a network provider. The service nodes refer to nodes with different functions and capable of providing service processing capacities of 4-7 layers for a network user. When being processed, a service transmitted in the network need to pass through a variety of different service nodes in turn according to a certain sequence.
As shown in FIG. 1, the network user (network user A as shown in FIG. 1) needs to customize a network service from a server device (which may be provided by the network provider) and then transmit a service according to the customized network service. Since the service nodes in the network are statically connected, all the services corresponding to a network user need to serially pass through all the service nodes in turn. In this way, in terms of the service nodes in the network, a large amount of unnecessary flow flows through the service nodes, resulting in large flow and low utilization rate of the service nodes.
In order to solve the problems of large flow pressure of the service nodes deployed in the network and low utilization rate of the service nodes, a technical solution of service chaining is provided. On the premise of meeting requirements of a service, the service nodes through which the service flows during transmission are flexibly controlled to avoid the service to flow through unnecessary service nodes when being transmitted in the network. A sequence formed by the service nodes corresponding to service transmission selected according to the service may be called a service chaining. The service chaining enables the network user to customize the service according to demand and pay according to the customized service. Internet engineering task force (IETF) proposes determining a transmission path of a service in a network based on a layer 3 virtual private network (L3VPN), specifically as follows: the transmission path of the service in the network is changed in a manner of manually configuring a provider edge (PE) router forwarding table, such that the service may pass through the set service nodes in sequence according to the configured router forwarding table. But in this manner, in terms of the network user customizing the service, the network user knows nothing about the specific locations of the service nodes in the network, and thus the determined transmission path of the service in the network could not represent the sequence of the service nodes in the network. Under this condition, when being transmitted, the service in the network still need to serially flow through the service nodes in turn, and thus the problems of large flow pressure and low utilization rate of the service nodes deployed in the network could not be well solved, and the efficiency of changing the transmission path of the service by means of manually configuring the PE router forwarding table is low.
To sum up, the general methods of determining the service transmission path could not well solve the problems of large flow pressure and low utilization rate of the service nodes deployed in the network.